无人工样本的SVM遥感图像分类方法

针对遥感图像监督分类方法需要人工提取训练样本的缺陷,提出一种模糊K均值聚类（FCM）提取训练样本、支持向量机（SVM）进行分类的方法。算法首先用FCM进行初步分类得到隶属度矩阵并判断每个样本的类别号;然后根据隶属度矩阵提取每类样本中密集程度较高的样本作为训练样本;最后用SVM对样本进行训练、再次分类。该方法克服了SVM算法需要人工样本的缺点,改善了传统非监督分类算法的性能,UCI标准数据库Iris数据和遥感数据样本的实验结果证明了该方法的可行性。     

一种集成式Beta过程最大间隔一类分类方法

一类分类是一种将目标类样本和其他所有的非目标类样本区分开的分类方法.传统的一类分类方法针对所有训练样本建立一个分类器,忽视了数据的内在结构,在样本分布复杂时,其分类性能会严重下降.为了提升复杂分布情况下的分类性能,该文提出一种集成式Beta过程最大间隔一类方法.该方法利用Dirichlet过程混合模型(DPM)对训练样本聚类,同时在每一个聚类学习一个Beta过程最大间隔一类分类器.通过多个分类器的集成,可以构造出一个描述能力更强的分类器,提升复杂分布下的分类效果.DPM聚类模型和Beta过程最大间隔一类分类器在同一个贝叶斯框架下联合优化,保证了每一个聚类样本的可分性.此外,在Beta过程最大间隔一类分类器中,加入了服从Beta过程先验分布的特征选择因子,从而可以降低特征冗余度以及提升分类效果.基于仿真数据、公共数据集和实测SAR图像数据的实验结果证明了所提方法的有效性.     

谱聚类划分随机森林算法在图像识别中的应用

随机森林是近些年发展起来的新集成学习算法,具有较好的分类准确率。针对该算法计算复杂度较高的不足,提出了一种基于谱聚类划分的随机森林算法。首先,利用聚类效果较好的谱聚类算法对原始样本集的每一类进行聚类处理。然后,在每一聚类簇中随机选取一个样本作为代表,组成新训练样本集合。最后,在新训练样本集上训练随机森林分类器。该算法通过谱聚类技术对原始样本进行了初步划分,将位置相近的多个样本用簇内的一个样本代表,较大程度地减少了训练样本的个数。在Corel Image图像识别数据集上的实验表明,算法可以用较少的分类时间达到较高的分类精度。     

一种集成式Beta过程最大间隔一类分类方法

一类分类是一种将目标类样本和其他所有的非目标类样本区分开的分类方法。传统的一类分类方法针对所有训练样本建立一个分类器,忽视了数据的内在结构,在样本分布复杂时,其分类性能会严重下降。为了提升复杂分布情况下的分类性能,该文提出一种集成式Beta过程最大间隔一类方法。该方法利用Dirichlet过程混合模型(DPM)对训练样本聚类,同时在每一个聚类学习一个Beta过程最大间隔一类分类器。通过多个分类器的集成,可以构造出一个描述能力更强的分类器,提升复杂分布下的分类效果。DPM聚类模型和Beta过程最大间隔一类分类器在同一个贝叶斯框架下联合优化,保证了每一个聚类样本的可分性。此外,在Beta过程最大间隔一类分类器中,加入了服从Beta过程先验分布的特征选择因子,从而可以降低特征冗余度以及提升分类效果。基于仿真数据、公共数据集和实测SAR图像数据的实验结果证明了所提方法的有效性。     

一种改进的支持向量机BS-SVM

提出了一种改进的SVM:BS-SVM,它先对训练样本进行分类,根据每个样本到模式类样本均值的距离,将训练样本分为三种:好样本、差样本、边界样本,然后用边界样本训练得到分类器.实验表明,BS-SVM相比SVM在分类正确率、分类速度以及使用的样本规模上都表现出了一定的优越性.     

基于优化Adaboost迭代过程的SVM集成算法

为提高Adaboost算法迭代过程中生成基分类器的分类精度以及简化整个集成学习系统的复杂度,文章提出了一种优化Adaboost迭代过程的SVM集成算法。该算法提出了一种在其迭代过程中加入样本选择和特征选择的集成方法。通过均值近邻算法对样本进行选择,并利用相对熵法进行特征选择,最后利用优化得到的特征样本子集对基分类器SVM进行训练,并用加权投票法融合各个SVM基分类器的决策结果进行最终判决。通过对UCI数据集的仿真结果表明,本算法与支持向量机集成算法相比,能够在更少的样本以及特征的基础上,实现较高的识别正确率。     

一种聚类模式下基于密度的改进KNN算法

KNN是基于实例的算法,训练样本的数量影响KNN的分类性能.合理的样本剪裁可以提高分类器的效率.提出了一种聚类条件下基于密度的KNN改进模型.首先使用聚类方法对训练集进行基于类别的选择,裁剪边缘样本以减少噪音;再基于类别密度对样本进行加权,改善k近邻选择时大类别、高密度训练样本的占优现象.试验结果表明,本文提出的改进KNN分类算法提高了KNN的分类效率.     

基于多尺度样本扩增的高光谱影像半监督分类

大量的训练样本可有效缓解模型过拟合,从而提高分类效果。在初始标记样本较少的情况下,开展借助不同尺度的同质区快速扩增大量高精度训练样本的实验,并利用初始标记样本和扩增样本训练支持向量机(Support Vector Machine, SVM)分类器,实现对高光谱数据的有效分类。该方法在Pavia University、Salinas和Indian Pines三种高光谱数据上均能获得大量高精度的训练样本,分类精度分别达到99%、99%和97%以上。实验结果表明,扩增的大量伪标签样本可以有效训练SVM分类器,提高分类效果。     

基于密度聚类与SVM的非法广播信号识别研究

针对非法广播信号的危害,以及传统人工检测效率低的问题,提出一种基于密度聚类与SVM的信号识别模型。首先,采用标准欧式距离对特征信号进行提取;其次,以聚类样本为基础,采用SVM分类器对信号分类;最后,以青海广播电视局中波台整点时刻前后300帧的数据为样本,以静音信号作为评价指标,对信号进行识别。结果表明,在正常信号中加入非法信号后,频谱中有少量的静音信号,且SVM训练时间和识别正确率都要优于传统算法。     

基于二层SVM多分类器的桩基缺陷诊断

 为了研究桩基缺陷的快速和准确的分类方法,依据支持向量机(SVM)理论,采用多层分类的方法,改进了一对一SVM多分类器结构,构建了二层一对一SVM多分类器模型,提出了二层多分类计算方法.与BP神经网络相比,二层SVM分类器的学习训练快捷,分类处理的实时性能好,对小样本测试环境的适应能力强,并且具有较好的分类准确率.该方法适合于分析训练样本数量少、分类精度要求高和分类输入输出变量较多的桩基缺陷诊断等多分类问题,对桩基多处缺陷识别的研究也具有重要支持.     

Cluster analysis of gene expression data based on self-splitting and merging competitive learning

Cluster analysis of gene expression data from a cDNA microarray is useful for identifying biologically relevant groups of genes. However, finding the natural clusters in the data and estimating the correct number of clusters are still two largely unsolved problems. In this paper, we propose a new clustering framework that is able to address both these problems. By using the one-prototype-take-one-cluster (OPTOC) competitive learning paradigm, the proposed algorithm can find natural clusters in the input data, and the clustering solution is not sensitive to initialization. In order to estimate the number of distinct clusters in the data, we propose a cluster splitting and merging strategy. We have applied the new algorithm to simulated gene expression data for which the correct distribution of genes over clusters is known a priori. The results show that the proposed algorithm can find natural clusters and give the correct number of clusters. The algorithm has also been tested on real gene expression changes during yeast cell cycle, for which the fundamental patterns of gene expression and assignment of genes to clusters are well understood from numerous previous studies. Comparative studies with several clustering algorithms illustrate the effectiveness of our method.     

基于模糊分类器及多层前馈神经网络混合结构的说话人辨认

提出了基于模糊超椭球聚类算法的说话人辨认新方法。该算法首先将某一类的训练数据分成若干子类, 对每一子类在其中心周围定义具有超椭球区域的模糊规则。实验表明,该系统可以较快的聚类速度取得与HMM 相当的识别效果。进一步的研究表明,基于模糊超椭球聚类算法的说话人辨认系统与传统的基于HMM的识别方法存在一个共同的缺点,即抗噪性能较差。为此,通过引入多层前馈神经网络(MLFNN)与模糊超椭球分类器构成混合模型,使系统的识别性能和抗噪能力显著提高。     

DBSCAN聚类改进算法与支持向量机结合的道路路锥识别研究

基于激光雷达传感器,提出了一种道路路锥识别方法。首先,在传统DBSCAN聚类算法的基础上改进算法搜寻核心点的方式,对雷达所采集的数据进行快速地分割、聚类。接着,分析类簇,对每帧数据的类簇进行特征采样并赋予标签值。最后,通过支持向量机(SVM)训练样本数据,利用网格化搜索与交叉验证法优化SVM参数,得到类簇分类模型,用于识别路锥。实验结果表明,改进后的DBSCAN算法计算效率有了显著提升,并且对点云的聚类更具有针对性。经过多次随机数据集检测,分类模型的准确率保持在93以上,实现了对路锥的有效识别。     

Image segmentation by clustering

This paper describes a procedure for segmenting imagery using digital methods and is based on a mathematical-pattern recognition model. The technique does not require training prototypes but operates in an "unsupervised" mode. The features most useful for the given image to be segmented are retained by the algorithm without human interaction, by rejecting those attributes which do not contribute to homogeneous clustering in N-dimensional vector space. The basic procedure is a K-means clustering algorithm which converges to a local minimum in the average squared intercluster distance for a specified number of clusters. The algorithm iterates on the number of clusters, evaluating the clustering based on a parameter of clustering quality. The parameter proposed is a product of between and within cluster scatter measures, which achieves a maximum value that is postulated to represent an intrinsic number of clusters in the data. At this value, feature rejection is implemented via a Bhattacharyya measure to make the image segments more homogeneous (thereby removing "noisy" features); and reclustering is performed. The resulting parameter of clustering fidelity is maximized with segmented imagery resulting in psychovisually pleasing and culturally logical image segments.     

一种基于动态遗传算法的聚类新方法

 如何确定聚类数目一直是聚类分析中的难点问题.为此本文提出了一种基于动态遗传算法的聚类新方法,该方法采用最大属性值范围划分法克服划分聚类算法对初始值的敏感性,并运用两阶段的动态选择和变异策略,使选择概率和变异率跟随种群的聚类数目一致性变化,先进行不同聚类数目的并行搜索,再获取最优的聚类中心.七组数据聚类实验证明该方法能够实现数据集最佳划分的自动全局搜索,同时搜索到最佳聚类数目和最佳聚类中心.     

An intelligent and knowledge‐based overlapping clustering protocol for wireless sensor networks

Overlapping is one of the topics in wireless sensor networks that is considered by researchers in the last decades. An appropriate overlapping management system can prolong network lifetime and decrease network recovery time. This paper proposes an intelligent and knowledge‐based overlapping clustering protocol for wireless sensor networks, called IKOCP. This protocol uses some of the intelligent and knowledge‐based systems to construct a robust overlapping strategy for sensor networks. The overall network is partitioned to several regions by a proposed multicriteria decision‐making controller to monitor both small‐scale and large‐scale areas. Each region is managed by a sink, where the whole network is managed by a base station. The sensor nodes are categorized by various clusters using the low‐energy adaptive clustering hierarchy (LEACH)‐improved protocol in a way that the value of p is defined by a proposed support vector machine–based mechanism. A proposed fuzzy system determines that noncluster heads associate with several clusters in order to manage overlapping conditions over the network. Cluster heads are changed into clusters in a period by a suggested utility function. Since network lifetime should be prolonged and network traffic should be alleviated, a data aggregation mechanism is proposed to transmit only crucial data packets from cluster heads to sinks. Cluster heads apply a weighted criteria matrix to perform an inner‐cluster routing for transmitting data packets to sinks. Simulation results demonstrate that the proposed protocol surpasses the existing methods in terms of the number of alive nodes, network lifetime, average time to recover, dead time of first node, and dead time of last node.     

基于数据选择模型的IB算法

针对数据对象自身模式特征明确程度的不同给IB（Information Bottleneck）方法数据分析带来的问题,定义一个“基于明确因素”的数据选择模型,使得IB方法可从数据集中选取模式特征较为明确的数据对象并对其进行模式分析,提出DSIB （Data Selection Information Bottleneck）算法.DSIB算法采用数据压缩过程中所产生的信息损失作为数据对象模式特征是否明确的判定条件,使用“边选择边学习”的顺序“抽取-合并”策略来优化DSIB目标函数.实验结果表明：随着数据选择标准的不断提高,DSIB算法在提高数据分析精度的同时所牺牲的召回率较小;与未做选择的数据分析算法相比,DSIB算法可更好地识别出数据中所固有的内在模式.     

基于类内差异和类间重叠的有效性函数

模糊C-均值(FCM)聚类算法的一个主要问题是需要事先确定聚类的数目,为此定义了类内差异度和类间重叠度来分别度量同一个聚类中数据的相似度和不同聚类间的分离程度,进而基于这两个度量提出一个新的有效性函数用于判定最佳聚类数目。实验结果表明,该有效性函数能有效地判定聚类数目,并且有较好的鲁棒性。     

Evolution-Based Tabu Search Approach to Automatic Clustering

Traditional clustering algorithms (e.g., the K-means algorithm and its variants) are used only for a fixed number of clusters. However, in many clustering applications, the actual number of clusters is unknown beforehand. The general solution to this type of a clustering problem is that one selects or defines a cluster validity index and performs a traditional clustering algorithm for all possible numbers of clusters in sequence to find the clustering with the best cluster validity. This is tedious and time-consuming work. To easily and effectively determine the optimal number of clusters and, at the same time, construct the clusters with good validity, we propose a framework of automatic clustering algorithms (called ETSAs) that do not require users to give each possible value of required parameters (including the number of clusters). ETSAs treat the number of clusters as a variable, and evolve it to an optimal number. Through experiments conducted on nine test data sets, we compared the ETSA with five traditional clustering algorithms. We demonstrate the superiority of the ETSA in finding the correct number of clusters while constructing clusters with good validity.     

An energy‐efficient clustering approach for collaborative data forwarding in heterogeneous MANET

Mobile ad hoc network consists of a group of mobile nodes that can communicate with each other without any infrastructure. Clustering of the mobile nodes ensures efficient use of available bandwidth and high network throughput. Various clustering schemes are developed to improve the energy efficiency and lifetime of the network. However, there is an increase in the energy consumption with the increase in the number of clusters for forwarding data. This paper presents an energy‐efficient clustering approach for collaborative data forwarding in mobile ad hoc network. The cluster head (CH) is selected based on the processing capability of the nodes and link connection metrics. The CH receives the data from the server and forwards the data to the member nodes at a corresponding data rate of the nodes. Data offloading technique manages the data traffic in the network. The CH rejoining approach enables load balancing in the network. The proposed clustering approach achieves a significant reduction in the energy consumption and data traffic and improvement in the throughput rate through stable routing.